AUTOMISED CALIBRATION OF EMPIRICAL LOSS - AND DEVIATION MODELS FOR COMPRESSOR BLADE ROWS

摘要

Within a project of German research co-operation AG Turbo, the operational behaviour of a 15-stage transonic compressor rig was experimentally and numerically investigated. The rig is geometrically scaled from an existing advanced heavy-duty gas turbine compressor with CDA-blading. The development and validation of a process for automatic calibration of empirical loss and deviation correlations implemented in a compressor map calculation program for multistage axial compressors (mean section performance calculation), based on existing compressor map measurement data is the target of this project. In Order to calibrate the empirical loss - and deviation models 1D-data as overall pressure ratio, mass flow and polytropic efficiency was extracted from the measurement data. Subsequently the coupling of an optimisation algorithm, the compressor map calculation program and the extracted measurement data had to be realised for the automised process of calibration as the second step. Due to different targets for the parameterised loss - and deviation models they were adjusted separately with different objectives as mass flow for the deviation model and as efficiency for the loss model. Following the cumulative calibration was carried out with a combined target of mass flow and efficiency deviations. The process for the automatic calibration of the empirical loss and deviation correlations implemented in a compressor map calculation program has been successfully developed and validated for different speed lines based on the measurement data of a 15-stage compressor rig.